Once again I've lost a period of work because it did not save. I think that it is down to taking too long and I need to save every 30 minutes or so ... I think. Anyway now to try and recreate what was a nice set of notes :(

So why does it take so long to rip a CD compared to a DVD? MakeMKV will process a DVD in several minutes, but does not provide the correct names for each chapter, while a CD is displayed with track names but which may take 20 times longer to copy for a tenth of the data. The external DVD drive takes a longer than the old internal CD drive so while I can rip 2 at a time, it's not twice as fast. Windows actually does it quicker, but only to WAV and I tend to use FLAC, and again it's an older CD drive which seems to help. But it's a Windows 10 machine that can't be upgraded ... yet another irritation that the think keeps complaining that I need to upgrade to a new machine when it does fine for the jobs I use it for ... such as a media player to get around the aggro from the LG Smart TV. Something else to document.

Anyway, I'd scanned all of the popular music and they are all up on Plex. Given the small number of CD's left that did not take long, but I was left with the classical music CD's. A set of 63 disks, each in books which was I think a fortnightly magazine program back in 2003/4. The Classic Composers on MusicBrainz and my local list The Classic Composers! The problem that comes up is that many of the disks bring up track data that is not in English, or even none at all as is the case on the windows box. So just where is the information coming from? In the past it would have been FreeDB, but I know that has gone, so what is replacing it today. Internet searches were as useless as usual, especially when trying to focus in on openSUSE related answers, so I resorted to DudkDuckGo's chat function to see if I got any better results. Llama 3.1 is apparently providing the results, and after a few questions that it got essentially got right it started to struggle with the openSUSE specific ones, giving duff directions a number of times before we hit the correct one. AI is another rant project but that will be later. That things have changed since Llama 3.1 was 'trained' is obvious and I don't believe any of them are currently intelligent enough to fix that training and when someone else asks ... go to the correct answer first.

The location I was looking for is under System Settings -> Hardware -> Multimedia and this provides a couple of useful pages. 'CDDB Retrieval' confirms that openSUSE uses both MusicBrainz and FreeDB although the second is actually accessing gnudb.org to provide the data. While there is a generic URL for the second, the site advises that using a registered one is better, and having set that up, the response times seem to have improved a little. The problem I had though was many of those results were in a foreign language. Page also identified the location of the cddb data cache which is set to a hidden directory in the home folder '.cddb'. Looking through that showed multiple copies some of which were English, against the duff one being served up. However a number of my disks did not get any results so just how does one add data somewhere.

The submission process for FreeDB pointed me to CUERipper and I have that on the Windows machine but I've not been able to get it working on Linux yet. At the same time I found a package called Picard which provides the same service for MusicBrainz. I've not got my head around it fully yet, but have enough working to be able to maje a good start at populating the collection metadata. I've been creating United Kingdom releases of each disk, some of which link to later American releases or track lists from other sources, and a few I've loaded up manually.

I've pulled my crib-sheet on Processing a disk on MusicBrainz to it's own page. I'll expand on that as I learn a few more of the tricks. I will also expand on my standards for the filing system in line with how Plex and other tools need things layed out. That dictates my own file naming conventions. It would be nice to have a DVDBrains sites for DVD's as even MakeMKV only puts numbers up when ripping a disk.

The MAJOR irritation with MusicBrains is the fact that some material is ONLY displayed in a foreign language such as Пётр Ильич Чайковский which windows struggles with in a file name. Yes Pyotr Ilyich Tchaikovsky is displayed sometimes, but not in the important areas such as the cddb file. This explains why I was struggling so much at first. The system needs to respect the language entered for a release and provide the right transliteration on any display. Or do we just have to create a separate composer entry in the same way as the same disk ID can have multiple releases. One to look at going forward, and I think I've caught up with were I was this morning ...

Add to translation list Николай Андреевич Римский‐Корсаков, Модест Петрович Мусоргский, Александр Порфирьевич Бородин, Արամ Խաչատրյան, Сергей Васильевич Рахманинов, Дмитрий Дмитриевич Шостакович